Hacker News Reader

GP is talking about something else, and arguably in agreement with your view: for a technology to come about and stick, you need a perfect storm: all the pieces being available and a source of continued demand for it. The former you need for an invention to be made; the latter you need for it to survive - to continue being made and to spread worldwide.

WRT. rocketry, this is a story of technology getting almost, but not quite there - it was all evolving at a rapid pace, until suddenly demand disappeared. It's not that demand for rockets wasn't there at all - just not at the price point which designs from the 80s/90s commanded What Elon did with SpaceX, was to focus on dropping the price. That involved revisiting the reusable boosters idea, which didn't pan out then because they didn't have to worry about money as much. It got perfected and productized by SpaceX now, because it was a road to cheaper launches - and they got them cheap enough to meet the existing demand (and create more of it).

Also for rocketry the established players didn't want to majorly reduce the cost because they had government monopolies. Innovators dilemma.

I used to watch a show called Connections. Things can exist in other fields that cross over and make other fields absolutely ignite with interesting possibilities.

SpaceX is taking existing tech and rebuilding it up from ground up principles. It is the method they are applying to all of his businesses. Get something working. Then go back and throw everything out until it stops working. Then put that thing back. Then throw out more. The idea is if you still work perfectly fine without something you did not need it. Following the optimization method of 'the most optimal thing is the thing that is not there.' It is what took Tesla from a bespoke 1-5 cars per year company to the capability to build thousands per week. It is a brutal painful process that works.

The idea you are specifying is called Muntzing: https://en.wikipedia.org/wiki/Muntzing - named after a CEO who supposedly made the poorest TVs on the market, but also the cheapest. Which left him with lots of money to spend on marketing those poor-quality TVs to people who they would not work for, who had to pay for their own return shipping.

The capability of building thousands of cars per week is quite old. You may be referring to EVs, but even in that category we have BYD that is knocking competitors out the park.

It's increasingly debatable that the process actually does work.

The tech necessary for steam engines was available by the late Roman empire

Archimedes even demonstrated the principle of heated water caising a sphere to rotate

But there was little use for it, without substantial investment in other areas. Horses for land transportand human rowers for ships were just cheaper and more practical

It would take another 1500 years before Watt found its use for a pump

> The tech necessary for steam engines

They had basic concepts but could not have produced a steam engine that did useful work.

The Archimedes device you mention (the aeolipile) was basically a kettle mounted on an axle that did indeed spin. But it had no way of building up pressure to the degree necessary to move something other than itself.

The Romans probably didn't have good enough materials tech to build vessels that could hold sufficient pressure to do useful work. Early steam engines could explode catastrophically and they were built by cultures that had centuries of cannon-making experience behind them. The Romans also lacked a compelling application that would have driven tech development and didn't have the iron/steel/coal industry required to build railways, etc.

It's telling that the first commercially viable steam engines were basically water pumps for getting water out of mines, and it didn't matter if they were massive and underpowered. It took another 60/70 years for the tech to improve enough to power vehicles.

I also doubt the Romans had tools that could have machined a piston with sufficient precision either.

FWIW, there was plenty of use for pumps earlier on - in mining. But muscle power, then water power, was sufficient.

Part of the steam engine's success was how it contributed to itself - it helped mine coal that would fuel it, and that would fuel furnaces in which steel was made, some of that steel going into... making more engines - and all of that moved around by means of steam power.

*Factorio vibes intensify*

You had to have enough pieces together: demand for coal and steel and moving them distances, sustained long enough to exhaust simple options, and become a problem the steam engine could solve. It was then that the steam engine turned this into self-reinforcing loop, and kickstarted the modern world.

That exponential character is what we are seeing with computers too. Better computers increase both the supply and the demand for even more.

Your discussion with DrF jogged my memory a bit. There was a discussion on this topic some time ago (I'm fuzzy on when) that went into detail about how there needed to be a specific use case and demand for this. I'm trying to remember the blog, but a search on HN yielded this: https://www.ageofinvention.xyz/p/age-of-invention-why-wasnt-...

I think this is the blog in question, and the most relevant part. If something else comes to me, I'll share it here.

The steam engine came about due to the spread of shipworms. Ships need copper shielding against that, thus mining deep, thus industrial revolution. First comes the demand.

Which came about by the merchant fleet of the British empire.

"The tech necessary for steam engines was available by the late Roman empire"

I read here on HN that this might be not really true. Because the romans did not heavily invest into high pressure chambers, aka cannons, like it was done in medival times. You don't just need a metal pot to build a steam engine. They must be able to hold much pressure and if they don't, they explode.

So the romans likely would have been able to make it work, but it would have required a significant investment - as they did not had experience in building pressure chambers.

> The tech necessary for steam engines was available by the late Roman empire

So long as you don't also add the "useful" qualification. If you want useful steam engines you things that didn't exist then. The iron alloys available to the Romans didn't allow for useful steam engines. Even the first steam engines were only useful in deep coal mines (where fuel was practically free) which the Romans didn't have.

Relevant: https://acoup.blog/2022/08/26/collections-why-no-roman-indus...

> The tech necessary for steam engines was available by the late Roman empire

IIRC I read somewhere it depended on the previous development of firearms, as it allowed metalworking to advance enough to produce metal tubes with very smooth and regular bores, that avoid leaks, jamming and uncontrolled explosions, to be fitted with pistons. The Roman empire had no firearms.

There were 100s of Newcomen engines throughout Europe which were mostly used to pump water. They were being deployed ~50 years before Watt improved the design.

I think we can say that a successful airplane was inevitable in the early 20th century because of the other work being done in the field at the time. Other designs were getting close, but didn't quite have the necessary combination of power to weight and lift to drag. The most important innovation from the Wrights was the realization that aerodynamic roll control was necessary; it's harder to say how many years would have been needed for someone else to try that.

That's not to say the same thing is true for any other invention or technological advance.

> Yes absolutely, environment makes a difference, but it also looks pretty clear that the raw tech can exist for something for a very, very long time without anyone putting the pieces together.

Perhaps invention is slightly different, but in the empirical sciences there's a bunch of stuff that occurred at roughly the same time:

* https://en.wikipedia.org/wiki/List_of_multiple_discoveries

* https://en.wikipedia.org/wiki/Multiple_discovery

Science and technology are closely related, as the latter builds on the former, and so it is often the case that sometimes a little bit of luck / timing determines who is "first", e.g.

* https://en.wikipedia.org/wiki/Elisha_Gray

I think there is also a bifurcation to be had on inventions which take billions of dollars into a business to R&D and tens or hundreds of millions of dollars per test vs inventions one mildly wealthy (or less) average individual can reasonably cover the investment of. The former is less about the invention making and more about the funding gathering.

The point is, the archetypal inventor is someone who has a unique flash of insight that only they could have had. Your example is of something different: that with enough funding and smart people working on a problem, great things can be achieved. Elon didn't personally have almost any of the insights, but he did supply the vision and funding.

He supplied a lighting rod for negativity.

The first few launches were very difficult. A NASA project with 10 crashes would have been destroyed by the press and congress.

The superpower of Elon is to shield the technical team from the negativity, pipe it to null and post "close but no ciggar".

NASA's very early launches were hardly more reliable. But there was always an overlap with ICBM development and national defence. So reliability improved quickly, and one of the goals of the space program was to showcase that reliability.

Once Apollo was running the success rate was 100%, with no boosters lost - an incredible achievement, not just of tech but of project management.

NASA of the day didn't need shielding, because it was already doing the job with no excuses.

Well, only if you ignore Apollo 1. Overall Apollo was incredibly successful, but not 100%.

At that time they didn't have TV news channels that can repeat the explosion 24 hour a day. Also the space race and cold war was a good shield, if someone complain s/he was accused of being an evil communist spy trying to destroy America.

I think in he previos programs they had unmanned missions, but once they reached Apollo they (always?) have to use manned missions because docking was not automated. Once you have persons inside the rockets they should not explode, so you must be extra careful with the design.

> because it was already doing the job with no excuses.

We will have to wait until someone makes another cheap reusable rocket to know how hard it is to design them without exploding a few of them. The move fast and break think applied to rocket science may be a brilliant strategy or a stupid excuse, but we will not know until someone else try. (I'm not counting the Space Shuttle. It was not reusable, it was expansively refurnishable after month of work.)

You know, on second thought: I don't recall Elon attracting all that much negativity before his recent escapades. Robert Downey Jr famously played Elon in Iron Man, and the audiences cheered. I think this is a case of "we were always at war with Eastasia".

I agree it's worse now. At that time the bad comments were like

fake quote> He is rich only because his father had an esmerald mine, and he can not even build rockets with 1960 technology.

Thanks - usually I understand what my downvotes are for, and stand by them. This instance would have left me puzzled without this explanation.

I would have thought that a leader is judged by their results, and regardless of personal feelings for Elon - one can't argue that he managed to build SpaceX where many others would have failed. Could NASA have built something just as good? Possibly, but history doesn't do hypotheticals.

[I'm not sure that I understand the first part of ypur comment, but just in case, I didn't downvote neither of ypur comments.]

> Yet it would be 23 years before someone came around again.

Problem here is that we don't know what could have happened if USGOV had properly funded NASA and the sciences instead of cutting it back to the bone and then subsidising SpaceX.

IMO, even more important than funding is that NASA rockets are now designed by Congressional committee. It’s hard to innovate when you’re pressured into reusing Shuttle components to keep the money flowing to specific contractors.

We certainly do know that while the Apollo program was still in progress NASA was developing the nuclear rockets that would make Mars missions feasible. Think weeks of travel time instead of months, and much larger payload fractions. If you want to talk about funding cuts, talk about Nixon. The Commercial Crew program came decades later after the failure of the Shuttle program, when it was revealed just how dangerous the they were.

It was reaveled how dangerous it was around 1986, after the Challenger disater. There were too many incentives to ignore it. Commercial crew would probably not come to be (at least not in its current form) if SpaceX wasn't around to prove that it's possible

It took another accident before anyone took the danger seriously enough to ground the Shuttle.

SpaceX has made a lot of real advancements while delivering real launches. You're thinking of Boeing.

Space in the US has always been extremely political. SpaceX is a more recent kind of politicisation which happens to be palatable to the current neoliberal orthodoxy, which follows the usual cycle of "break something the government runs by defunding it and interfering in its management, privatise it, then crow about how inefficient big government is."

If it had been left to the engineers in 1970s NASA and not to anti-science Republican cranks and crooks like Nixon and Proxmire, we would absolutely would have had a moon base and Mars landings long before now.

There is a vast gulf of time between the 1970s and SpaceX. The space shuttle alone, fully NASA-run, was a (beautiful-looking) failure long before SpaceX came along. And it was fully government-designed and government-run. It was also astonishingly expensive.

Good things can happen in the private sector or the public sector. SpaceX is a good thing happening in the private sector. No one's brain has to go running to Republican-bashing and history compression to counterbalance that to avoid thinking new thoughts.

The space shuttle had government specified requirements (a single vehicle to do everything including capturing soviet satellites and ferrying people into orbit)

It had to do everything because the business case for it (that it would have sufficient ROI) required it. Even then, the business case was basically fraudulent, and the reality was even worse than the critics like Mondale were saying.

No engineer that works for SpaceX, with SpaceX, or is closely connected to engineers who work for SpaceX would agree with your assessment that SpaceX's success is due to:

> "break[ing] something the government runs by defunding it and interfering in its management, privatise it, then crow about how inefficient big government is."

It is possible that this will occur going forward now that Elon is a de facto government official, though. Also, the engineers in 1970s NASA are what created the Space Shuttle.

I have no idea why you think NASA would have gotten to the moon starting in the 70's. Every single dollar was being spent flying the Shuttle and setting up the ISS during the 80s-2010's.

>Problem here is that we don't know what could have happened if USGOV had properly funded NASA and the sciences instead of cutting it back to the bone and then subsidising SpaceX.

Uh, SpaceX didn't come into being until 2002. And the Space Shuttle was not in any way cheap, nor a pile of things we did with it. Nor did the US Government in any way subsidize SpaceX, anymore then it "subsidizes" paper manufacturers by... ordering paper for its printers. It contracted with SpaceX for commercial services at a fixed price, and has saved billions and billions of dollars as a result. You seem a touch confused on timelines here.

And my entire point was that whatever could have been done, it wasn't. As I said, the government absolutely could have pursued other far fundamentally better concepts at the end of Apollo. Or it could have done commercial way earlier, getting out of the launch business entirely and working to switch over to a competitive private market 30-40 years earlier. What actually happened is that without the focus and level of discipline provided by a big ambitious goal and national spirit as a guiding star, more typical political incentives crept in and rapidly distorted the space program. Which resulted in enormous waste and stagnation as well as killing a lot of incredible people for no reason.

But regardless of how it happened, again the point is that it wasn't the scientific and tech environment that created a small window of a few years where inevitably practical economic mass launch designs would come about. It still took the right vision and right spark.

Technically NASA contracted with SpaceX (and several other companies as well, such as Boeing) to develop crew–rated versions of their rockets. That was explicitly a development contract where NASA paid a couple of billion up front to entice those companies to develop a crewed craft capable of visiting the ISS. Later they bought rides to and from the ISS from SpaceX and Boeing. SpaceX used that money and some of their own to go from the Falcon 1 to the Falcon 9. Since then NASA has bought flights on Falcon 9 rockets for launching satellites and space probes in addition to crewed flights to the ISS.

So yea, definitely a prudent investment. But also some subsidies in there, and a few billion wasted on Boeing. And even more billions wasted on SLS too.

>Technically NASA contracted with SpaceX (and several other companies as well, such as Boeing) to develop crew–rated versions of their rockets

No, the initial NASA contract was for cargo delivery via the Commercial Orbital Transportation Services (COTS) program, which dated to 2006. CRS and CC came significantly afterwards. Momentum did build but it was in fits and starts. It's a pretty fascinating bit of history in turns of twists and turns on both the political and technical sides, I highly recommend taking a look at the book Liftoff if you're at all curious. Like, even if it was selfish motivations it's worth noting that Commercial Crew probably would not have happened in 2010 without the support of Boeing itself.

>But also some subsidies in there, and a few billion wasted on Boeing.

Eh, I wouldn't fully agree with this sentiment honestly. The future we've arrived at was definitely not clear when the contracts were signed, and shit happens. If you're really serious about developing a new capability and place a few fully independent investment bets and some pay off handsomely while others are a bust, I wouldn't call it a waste because if you just couldn't know which would work and which wouldn't. That's just how it works. It's worth noting here too that Boeing has lost enormous amounts of money on its failures with Starliner. NASA has held them to the fixed price aspect of the contract. I suppose an argument could be made that NASA was more generous with milestone payments then they might have been, but at the same time that ultimately was Boeing shooting itself in the foot and NASA giving them rope because getting some money earlier didn't change the total pot at all. So if Boeing rushed, well that came back to bite them very hard.

>And even more billions wasted on SLS too.

That's pure old space pork and a totally separate discussion beyond being another example of how decades can flow by on garbage.

> If you're really serious about developing a new capability and place a few fully independent investment bets and some pay off handsomely while others are a bust, I wouldn't call it a waste because if you just couldn't know which would work and which wouldn't.

Sure, that’s true. But my point is that those initial investments were a subsidy. They weren’t buying a product that already existed, they were paying someone to develop a product so that they could buy the product once it existed. It definitely paid off, even counting the money spent on Starliner.

> That's pure old space pork and a totally separate discussion beyond being another example of how decades can flow by on garbage.

Agreed. I am just saying that the left hand doesn’t know what the right hand is doing. At this point we’d be better off cancelling SLS and doing our next moon mission by launching smaller craft into Earth orbit on Falcon Heavy and docking them together. Like Apollo but with several launches spread out over a week instead of a single giant rocket. We could even assemble a cycler that way.

> we don't know what could have happened if USGOV had properly funded NASA

Sure we do, just watch For All Mankind :-) https://en.wikipedia.org/wiki/For_All_Mankind_(TV_series)

A more interesting question is what could have happened if NASA had used much more of its funding on practical research (including rocketry research), instead of wasting a huge amount of it on sending people up into space for no good reason over the past 50 years.

Currently, almost half of NASAs budget is spent on manned space flight.

> You can find endless examples once you start looking

My favorite examples are gunpowder, the printing press.

Both were known by the Chinese and Koreans for thousand years before the Europeans. However, it was the Europeans that began using gunpowder in handheld guns (Chinese used it for fireworks) and combined the printing press with metallic moveable types and oily ink, making it far more efficient.

Also: how much James Watt improved the already existing steam engine.

> Skipping forward, the basic controls of fully automated vertical landing were directly demo'd in a real flying 1/3 scale test bed (though important to note not an orbital one) with the DC-X in 1993.

Why is it important to note that it was not orbital?

>Why is it important to note that it was not orbital?

Orbital is what actually provides any significant economic value. But the nature of the Rocket Equation, effects with atmosphere, and how those interact with the limits of our material science means that there is an absolutely enormous gulf between a sounding rocket and getting to orbital velocity and back again. All of that is what makes it so hard but it's also required to be of any real practical use. So the DC-X was a noteworthy historical milestone/demo and test bed but going from there to something like the Falcon 9 let alone Starship takes an tremendous amount of further innovation and engineering (and investment of course). The new engine work alone SpaceX has done is a hufe deal and is what defines the envelope for the rest of the system.

An orbital craft is moving at hypersonic speeds, which is a especially tricky when going backwards. Things moving through fluids tend to want to flip over so that the part with the highest drag is at the rear, so that drag is minimized. Meanwhile the engine bay of a rocket is not generally shaped to have low drag when pushing through the air.

Wouldn't any spacecraft designed to return to Earth non-destructively, regardless of whether it was going to land vertically or land like an airplane or land with a parachute, first slow down to subsonic speeds before getting to the phase of the flight where the landing system is invoked?

It slows down by burning stuff and making hot stuff go out one end. That end is the one worse for air drag, because it has all the bits where the hot stuff comes out.

No. Go watch a Falcon 9 booster recovery <https://www.youtube.com/watch?v=w-_90o1KLkM> on Youtube; you’ll see that when it starts its landing burn it is ~5 miles up and flying engines first towards the ground at ~1300 miles per hour.

> the basic controls of fully automated vertical landing were directly demo'd in a real flying 1/3 scale test bed (though important to note not an orbital one) with the DC-X in 1993.

Surveyor 1 was the first automated vertical soft-landing rocket AFAIK, in 1966. DC-X was the first using turbo-pump engines.

This is why I say people that say private capitalism is what moves the world tech forward are wrong. From Rockets, Nuclear, Robots, Internet, Cars, Planes, Semi Conductors, Satellites all these came about from government investment in loss making technology or by heavily subsidizing the production for war until it reached profitability.

I mean, both have a big role to play. You mention rocket tech, which is a perfect example of the strengths and weaknesses of public and private enterprise.

Public enterprise can be great because it enables societies to leverage huge amounts of capital (at one point, the US was spending 4% of GDP on the Apollo program each year--the equivalent of $1 Trillion/year today) to accomplish goals that are beyond the capacity of private industry at the time. It suffers from being totally divorced from the normal constraints of economy, and so tends to way overspend (4% of GDP is frankly insane) and to get stuck at local maxima (see SLS).

Private enterprise is basically the opposite. It can never match the government in terms of spending big to pull tech forward by decades--the profit margin doesn't really work that way. But it lavishly rewards the efficient allocation of capital, which (so long as there is meaningful competition) creates a drive to constantly improve tech. It's painful to think about where rocket tech would be without the recent advances born of private capital!

> Elon Musk and SpaceX ultimately invented

Why do you keep giving engineering credit to a Paypal Billionaire? By that logic, how much invention did Bezos do with BlueOrigin? I guess none other than opening a check book?

> By that logic, how much invention did Bezos do with BlueOrigin?

Why didn't BlueOrigin surpass SpaceX? Is it only a funding issue?

wait, do you think the reason is that elon musk personally helped invent stuff?

No, but if it's not a difference of "Musk(programmer) vs Bezos(businessman)" I would like to know why other companies or leaders with a lot of money didn't have such success. Is it only a size of bank account? To me it looks like a ceo or a leader of company has more correlation with success than "just a lot of money".

My take - a CEO that is technical enough to decide on technical matters in inevitable arguments over what direction to choose is better for tech company than a business CEO that only looks for "what is cheaper in the short term".

Bezos made some pretty technologically prescient decisions too, the obvious example being the API mandate, which set the ball rolling for AWS. I think leadership can be a factor, but to me what you're doing is a massive simplification. In any case, I don't think being a programmer would make me more qualified than anybody else to make good decisions for a rocket building company.

> a CEO that is technical enough to decide on technical matters in inevitable arguments over what direction to choose is better for tech company than a business CEO that only looks for "what is cheaper in the short term".

You do know Bezos has degrees in electrical engineering and computer science from Princeton unlike Musk, right?

I have a computer science degree too, but looks like it doesn't make me more successful. So I'm still asking what is the difference that made SpaceX more successful than BlueOrigin.

Well to start with you should know Musk has never been a "Programmer"

https://www.businessinsider.com/dogecoin-creator-says-elon-m...

The evidence is:

> He asked, 'How do I run this Python script?'" Palmer said, per Crikey.

That is exactly what a programmer would ask, python scripts are not self contained they have a lot of dependencies, also typically you need to invoke them with a lot of parameters.

In addition not all programmers are python programmers, getting some commands you need in order to get started is just run of the mill when you get code from someone.

He does not have a clue: https://youtu.be/cZslebJEZbE

This is the same argument people used against the Wright brothers that you're now throwing at Elon Musk (in a thread maligning the Wright brothers claim to fame)...Langley was a classically trained mathematician with associations to Harvard and the Naval Academy. He was one of the most sought after and decorated engineer's and astrologers (invented his own time system) and was given exclusive funding by the Smithsonian to solve flight. He lost to the Bicycle repair shop guys that hand rolled their own equipment on sandbars in North Carolina (who didn't have advanced degree's in engineering).

You can do a lot with 38 Billion dollars of US Government money. It's a very Socialist thing to do: https://www.washingtonpost.com/technology/interactive/2025/e...

Both Musk and Bezos are businessmen. There is still competition among businessmen to do business the best way - directing the right people to work on the right things. But don't confuse this with actually making the products themselves. One can be a good businessman or a bad businessman, but this has nothing to do with engineering.

(Arguably, the best thing Musk did for SpaceX as a businessman was to delegate all the business to someone else. He tends to ruin companies where he does not.)

Yea, there are hundreds or even thousands of billionaires. Only one of them has developed a successful orbital launch company. That’s not exactly a high base rate.

And how many have tried? Not more than 5 would be my guess. We do not have enough data to make a conclusion.

The ones who didn’t try knew that they couldn’t do it. We do have enough data to conclude that money alone is not enough.

Musk happened to hire better engineers and programmers I guess

Musk is likely a worse engineer than Bezos and less technical. So that is probably not the reason. What Musk instead is is an excellent salesman. But maybe he was just lucky. Whatever it was it was not engeeeing which is a big weakness of Musk.

He seems to be the champion of using stainless steel for the rocket, which was a game changer.

Not a novel idea. The Atlas Family of ICMBs, the Thor/Delta Rockets, The Space Shuttle External tank, The R7 Russian rockets and the N1 Moon Rocket. They all did.

Musk's actually successful rocket is not made from steel. It has yet to be demonstrated that Starship is rapidly reusable, let alone that steel is significantly superior than other materials for this purpose.